TheSilverInstitute revises data 5 years later = normal business?

Pirocco

Pirocco

Well-Known Member
I came across an older PDF showing the supply/demand data from period 2000-2009:
http://fs.usda.gov/Internet/FSE_DOCUMENTS/stelprdb5336488.pdf
2000 2001 2002 2003 2004 2005 2006 2007 2008 2009
Supply
Mine Production 591.0 606.2 593.9 596.6 613.0 636.8 640.9 664.4 684.7 709.6
Net Government Sales 60.3 63.0 59.2 88.7 61.9 65.9 78.2 42.5 27.6 13.7
Old Silver Scrap 180.7 182.7 187.5 183.9 183.7 186.0 188.0 181.8 176.0 165.7
Producer Hedging -- 18.9 -- -- 9.6 27.6 -- -- -- --
Implied Net Disinvestment 87.1 -- 12.6 -- -- -- -- -- -- --
Total Supply 919.1 870.9 853.1 869.3 868.2 916.3 907.2 888.7 888.3 889.0
Demand
Fabrication
Industrial Applications 374.2 335.6 340.1 350.8 367.6 407.0 427.0 456.1 443.4 352.2
Photography 218.3 213.1 204.3 192.9 178.8 160.3 142.4 124.8 104.9 82.9
Jewelry 170.6 174.3 168.9 179.2 174.8 173.8 166.3 163.5 158.3 156.6
Silverware 96.4 106.1 83.5 83.9 67.2 67.5 61.0 58.4 56.9 59.5
Coins & Medals 32.1 30.5 31.6 35.7 42.4 40.0 39.8 39.7 65.2 78.7
Total Fabrication 891.7 859.4 828.3 842.4 830.8 848.7 836.4 842.5 828.6 729.8
Producer De-Hedging 27.4 -- 24.8 20.9 -- -- 6.8 24.2 11.6 22.3
Implied Net Investment -- 11.4 -- 6.0 37.4 67.6 64.0 22.0 48.2 136.9
Total Demand 919.1 870.9 853.1 869.3 868.2 916.3 907.2 888.7 888.3 889.0
Silver Price 4.953 4.370 4.599 4.879 6.658 7.312 11.549 13.384 14.989 14.674
Click to expand...
Compare the 2003+ data with the current:
2003 2004 2005 2006 2007 2008 2009 2010 2011 2012
Supply
Mine Production 597.2 613.6 637.2 641.3 666.1 683.0 713.6 752.7 757.0 787.0
Net Government Sales 88.7 61.9 65.9 78.5 42.5 30.5 15.6 44.2 12.0 7.4
Old Silver Scrap 196.0 198.7 202.7 206.2 203.0 200.8 199.8 228.8 258.1 253.9
Producer Hedging 45.9 50.4 12.2
Implied Net Disinvestment 7.8 1.4
Total Supply 889.8 874.1 951.7 926.0 913.0 914.3 929.1 1,076.2 1,039.4 1,048.3
Demand
Fabrication
Industrial Applications 368.4 389.7 430.3 453.0 486.2 490.9 403.6 500.7 487.8 465.9
Photography 192.9 178.8 160.3 142.2 117.6 101.3 79.3 72.1 66.1 57.8
Jewelry 186.8 187.6 188.4 176.5 183.8 179.1 178.7 192.8 186.5 185.6
Silverware 85.1 68.3 69.6 63.4 61.5 59.8 55.0 52.8 48.3 44.9
Coins & Medals 35.7 42.4 40.0 39.8 39.7 65.3 78.8 99.4 118.3 92.7
Total Fabrication 868.8 866.7 888.6 874.9 888.9 896.4 795.4 917.9 907.1 846.8
Producer De-Hedging 21.0 2.0 11.6 24.1 8.7 17.4 41.5
Implied Net Investment 5.4 63.1 39.5 9.3 116.3 158.3 132.3 160.0
Total Demand 889.8 874.1 951.7 926.0 913.0 914.3 929.1 1,076.2 1,039.4 1,048.3
Silver Price 4.879 6.658 7.312 11.549 13.384 14.989 14.674 20.193 35.119 31.150
Click to expand...
The older pdf dates from 2010, and talks about 2009.
Look at the huge amount changed figures, even back in 2004, alike they adjusted 2004 numbers 5 years later. And the changes aren't exactly minor, take for ex. the underlined one, Implied Net Investment for 2005 was given as 67.6 Moz while 5 years later it was revised to 63.1 Moz, which is 6.7% less.
And that same Implied Net Investment the next year 2006, 64 Moz revised 39.5 Moz, which is 38% less.
And there are dozens others examples. Look at the entire scrap recycling, about all revised up 10% and more.
Let's now put this in the data acquisition situation, TheSilverInstitutes retrieves the figures from some UK based company, GMFS Limited.
So those guys there regulary phone TSI 'hey this and that figure from 5 years ago was wrong, please make it...'?
I don't know lol, but is a big amount revisions 5 years later, 'normal business' for data acquisition?
 
Yes it is, I've seen GFMS revised figure constantly. I think it has to do with the fact that not all countries release customs data etc on a timely basis, so if they waited for all data to be in their reports would be so out of date. So I think they estimate some figures.

Thing is they don't declare this. I've seen figures change from many quarters ago although it tends to settle down after a year or so.
 
bron suchecki said:
Yes it is, I've seen GFMS revised figure constantly. I think it has to do with the fact that not all countries release customs data etc on a timely basis, so if they waited for all data to be in their reports would be so out of date. So I think they estimate some figures.

Thing is they don't declare this. I've seen figures change from many quarters ago although it tends to settle down after a year or so.
Click to expand...
I can understand that some can be delayed for various reasons, and that waiting to publish demand/supply is no option.
But did you notice that the current supply&demand data proves that somewhere AFTER the 2010 PDF creation, so between 2010 and 2012, they revised data from 2003?
That's 7-9 years later. I find it hard to imagine a situation that causes data to be delayed for 7-9 years.
 
Eureka Moments said:
I reckon it proves what a waste of time it can be researching and analysing such data.
Click to expand...
Yes, buying silver at any price, that's the least waste of time, thus the way to go.
 
And since today the shifting last 10 years SI demand/supply page is replaced by a couple years (2011 and 2012) with just a few figures inside blabla sentences. No reason given for ceasing to publish all the figures in a table.
 
Eureka Moments said:
I reckon it proves what a waste of time it can be researching and analysing such data.
Click to expand...

^
+1000!

Don't tell Pirocco this, otherwise he won't have anything to do all day :P
 
leo25 said:
Eureka Moments said:
I reckon it proves what a waste of time it can be researching and analysing such data.
Click to expand...

^
+1000!
Don't tell Pirocco this, otherwise he won't have anything to do all day :P
Click to expand...
Your criticism applies to your own post. :P
 
Eureka Moments said:
tonyz92 said:
Eureka Moments said:
I reckon it proves what a waste of time it can be researching and analysing such data.
Click to expand...

Says the broke stacker
Click to expand...

I shall politely ask why you think Im a broke stacker?
Click to expand...
A possible reason for the thought could be that those that don't know figures about the silver market, have a higher chance to end up with a part of their dollars going to those that do know figures.
Just a polite guess.
 
An addition, because I started to understand an element better.

The "Implied Net DisInvestment" at the supply side, and the "Implied Net Investment" at the demand side, are figures derived from all others (hence the 'Implied'), any changes on the others accumulate together there.
TheSilverInstitutes data source company GMFS Limited assumes any difference between supply data and demand data they find, as investment-related.
For example, if they gather sales data totaling to 100 Moz, and gather purchases data totaling to 150 Moz, then they assume a Net Disinvestment of 50 Moz (so a -50 Moz at the the supply side under 'Net disinvesment'.
Another vice versa example, if they gather sales data totaling to 150 Moz, and gather purchases data totaling to 100 Moz, then they assume a Net Investment of 50 Moz at the demand side under 'Net investment'.
So any degree of data error (including absence of data - data gaps) is treated as investment-related.
For ex, if some industrial company buys 10 Moz silver for solar panels, but doesn't report it, while a silver mine does report the production of those 10 Moz, then GMFS Limited would add this 10 Moz to Net Investment on the demand side.
This likely explains the data revisions, although it's still weird to see it happen 5 years later. Hey we found a 5 years old sales document! :D

<year> <amount Moz>

>>> 2010???? reported
Implied Net Investment (negative = Disinvestment)
2000 -87.1
2001 11.4
2002 -12.6
2003 6.0
2004 37.4
2005 67.6
2006 64.0
2007 22.0
2008 48.2
2009 136.9

>>> 20130526 reported
2003 -7.8
2004 5.4
2005 63.1
2006 39.5
2007 1.4
2008 9.3
2009 116.3
2010 158.3
2011 132.3
2012 160.0

>>> overlapping years old > new table
2003 6.0 > -7.8 = -13.8
2004 37.4 > 5.4 = -32
2005 67.6 > 63.1 = -4.5
2006 64.0 > 39.5 = -24.5
2007 22.0 > 1.4 = -20.6
2008 48.2 > 9.3 = -38.9
2009 136.9 > 116.3 = -20.6

As one can see in the comparison table: ALL revisions are drops.
In the light of aboves explanation of Net (Dis)Investment and its examples, this means that GMFS Limited received/found purchases data for supply-demand differences that they previously assumed as investment.
But it's important to realize that this 'Implied Investment' can be just lack of data, and aboves 2003-2009 revisions prove that the amount silver purchases 'discovered later' total to 154.9 Moz. That's a substantial amount silver, 15,5% of a 1 year supply/demand of 1000 Moz. It's almost half the amount that is accumulated in the safes of IShares Silver Trust.
 
What I don't understand is why GMFS Limited doesn't just allow supply and demand to be inequal. That would make clear which part of the traded silver IS (dis)investment, and which part CAN be any of the categories.
Then they could revise data later on, without having to change the (dis)investment category when the data revision doesn't fall under it.
This way they are making (dis)investment ambiguous, and since the amounts silver that are involved in the revisions over the 7 years period total to 15% of world annual supply, it matters.

The current physically backed ETF stocks, as shown in my 'Silver info" topic total to 526.5 Moz. It's possible that I lack some, but I searched alot, so I should have all big/medium/smaller ones, and I doubt that eventual unknown ETF's would change this picture alot. If some1 knows one I don't, say it.

The 2004-2012 Net Investment totals to 682.8 Moz. That's 156.3 Moz more than current ETF stocks. Most of it was stockpiled over 2009-2012 so quite 'vulnerable' to revision.
2009 can give us a clue about this 'vulnerability' to revisions: originally it was 136.9 Moz, and later on it was revised to 116.3 Moz. The average revision over the 2003-2009 period amounted to 22.13 Moz.
So if we can assume this same 'degree of error', we should expect a 2010-2012 downwards revision of 3x22.13=66.39 Moz.
So our 2004-2012 Net Investment total of 682.8 Moz should then be revised down to 616.41 Moz. That's 89.91 Moz more than current ETF stocks. This must then be the amount silver bought outside ETF's, Coins & Medals and other categories. I wonder whether or not non-industrial1000ounce bars (so 1 ounce bars etc) fall under Coins&Medals.
 
It didn't catch my eye during the writing of my previous post, but now I noticed a very remarkable/weird coincidence in the data.

First is year, second is amount Moz supply or demand that year.
RATE is average Moz per year over the period.
Implied Net Investment (positive means it's [DEMAND], negative means it's [SUPPLY])
2000 -87.1
2001 11.4
2002 -12.6
2003 -7.8
TOTAL2000-2003 -96.1 RATE -24.025/YEAR
2004 5.4
2005 63.1
2006 39.5
2007 -1.4
2008 9.3
2009 116.3
2010 158.3 -
2011 132.3
2012 160.0
2013
2014
TOTAL2004-2012 682.8 RATE 75.867/YEAR
Click to expand...
versus
Coins & Medals [DEMAND]
2000 32.1
2001 30.5
2002 31.6
2003 35.7
TOTAL2000-2003 129.9 RATE 32.475/YEAR
2004 42.4
2005 40.0
2006 39.8
2007 39.7
2008 65.3
2009 78.8
2010 99.4
2011 118.3
2012 92.7
2013
2014
TOTAL2004-2012 616.4 RATE 68.489/YEAR
Click to expand...
See the two bold 2004-2012 totals?
Implied Net Investment totals to 682.8 Moz.
Coins & Medals totals to 616.4 Moz.
I already found it weird that this Net Implied Investment, which likely reflects (what else?) all the physically backed Exchange Traded Funds stocks of 1000 ouncer bars + some other non-fabrication stocks of similar bars, sat so close to the Coins & Medals total.
But then, as a step towards a new method to predict silver future relative price trend, I decided to do now what research company GMFS Limited will very likely do in the future: revising the Net Implied Investment numbers of 2010-2012.
To do so, I just calculated the average annual Moz of their past revisions.

Here again the revision data (see post #12):
2003 6.0 > -7.8 = -13.8
2004 37.4 > 5.4 = -32
2005 67.6 > 63.1 = -4.5
2006 64.0 > 39.5 = -24.5
2007 22.0 > 1.4 = -20.6
2008 48.2 > 9.3 = -38.9
2009 136.9 > 116.3 = -20.6
-13.8-32-4.5-24.5-20.6-38.9-20.6 gives -154.9 Moz, divided by 7 years gives an average -22.12857 Moz revision of the Implied Net Investment each year.
So what did I do: I already revised the 2010-2012 Net Implied Investment totals now, based on this average of their past corrections. 2010-2012 is 3 years so -22.12857 Moz x 3 = -66.38571 Moz.
Then aboves TOTAL2004-2012 682.8 Moz - 66.38571 Moz = 616.41429 Moz.
Guess what: it's exactly the Coins & Medals TOTAL2004-2012 616.4

There are now 2 explanations:
1) This is a crazy coincidence - crazy, because in the end, all used data is derived from averaging a period of 8 years, so surely no 'hit and run' coincidence - still, it is "POSSIBLE".
2) GMFS Limited / The Silver Institute, or anyone able to set/alter the published supply/demand data, is "throttling" the Implied Net Investment supply|demand to the Coins & Medals demand, over a multiyears time span. In short: manipulation. Manipulation has an objective, so the next question is then which. If one sets number Y to number X, then he's cheating Y. All data of the SilverInstitutes demand/supply is verifyable along other ways (Mint sales figures, mine production figures, and so on). The sole category that is hard to verify is exactly this 'Implied Net (Dis) Investment'. Because it reflects the mismatch between supply and demand figures, or the 'error' / data incompleteness. And there is a deliberate/chosen part in this error, as in my previous post #13 is explained. Why doesn't GMFS Limited use the published ETF stocks data? The serial numbers of the bars are also published, and the vaults contents are regularly inspected by agencies. So GMFS Limited has no 'publishable' reason to not trust the ETF stock figures, something else would be considered as a 'major scandal'. So IF this "throttling" indeed happens, then something is "horribly wrong" with the original number Y, in other words: the "real" Y would result in major contradictions with other data. In the extreme case (margin value) of Y being zero, then all those Exchange Traded Fund stocks, cannot exist.
Funny enough, this would also give an answer on the question why gold ETF's sold half their stocks earlier last year, while silver ETF's remained unchanged, yet the silver price dropping 30% anyway.

I'm no fan of both explanations, so I leave it open here.
Also, things are like they are, if everybody believes false data, then the market/price is like the false data is true, whether you like it or not.
 
The 2013 figures were just released.
https://www.silverinstitute.org/site/supply-demand/
The layout changed. They got rid of that "Implied".
They even changed Producer Hedging to Net Hedging Supply, combining supply OR demand side entry into one, just like I did in my tables.
They must have read my criticism on it here haha.
Yay have again some figures to crack in the next days! :D
And revisions - will be interesting to see if my 'pre-revision' that resulted in that weird exact same total for Implied Net Investment and Coins&Medals is confirmed.
Although revision may happen in future years too, upto 5 years later, and maybe longer because I lacked the data to see.

....

Well screw, they changed Coins & Medals to Coins & Bars and the figures are totally different.
And some other differences too, making comparison with past data much harder or impossible.
While 2013 was a pivot year, and comparison would have been very interesting.
Gonna be alot work to dissect what they did.
 
An update on this, we have now seen 5 years accumulated corrections, so time for a re-evaluation.

Implied Net Disinvestment
Positive figure = disinvestment so bullion sold from existing stocks ("destockpiled").
Negative figure = investment so bullion stockpiled.

2000-2009 data released in 2010 > data released in 2014
2000 87.1 > 94.7 = +7.6 Moz
2001 -11.4 > 1.9 = +13.3 Moz
2002 12.6 > 24.9 = +12.3 Moz
2003 -6.0 > 7.8 = +13.8 Moz
-----> average correction = 11.75 Moz
2004 -37.4 > 67.9 = +105.3 Moz
2005 -67.6 > 0.7 = +68.3 Moz
2006 -64.0 > 24.9 = +88.9 Moz
2007 -22.0 > 53.5 = +75.5 Moz
2008 -48.2 > 48.1 = +96.3 Moz
2009 -136.9 > -61.9 = +75 Moz
-----> average correction = 85 Moz

The total supply/demand average was over 2000-2003 878.1 Moz (figures released in 2010)
The total supply/demand average was over 2004-2009 907.7 Moz (figures released in 2014)
That's a mere 3.4% increase.
But the adjustments increased 623% (they 7-folded).
And think of this: every year Silver Institute and others make conclusions based on these figures.
Which after years turn out to have been completely wrong. In nearly all cases even from net investment to net DISinvestment. 2009, the last year of this time frame being the exception, but also least revised so over X years it might end as just that again.
So far all those 'conclusions'.
There is something seriously fishy with that Thomson Reuters data.
They abandoned the net investment/disinvestment since this year.
One may wonder why so... late, and why this supply/demand class existed in the first place, falsely suggesting investment while it was disinvestment.

The most recent figures are these:
2000 94.7
2001 1.9
2002 24.9
2003 7.8
2004 67.9
2005 0.7
2006 24.9
2007 53.5
2008 48.1
2009 -61.9
2010 -85.3
2011 -69.3
2012 -198.3
2013 -99.7
What will they report in 2018 for years 2009-2013?
Again all positive, ie move from net investment to net disinvestment?
 
Thomson Reuters just released new gold and silver market data to World Gold Council and SilverInstitute.
Remarkable (read: big) updates on existing 2005-2013 years:
GOLD (only year 2013-2014 changes)
- Net Government Sales 2013 has now increased to 625.5 tonnes. In 2014Q2 report it was already increased from the original 386.6 tonnes to 409.3 tonnes. So the total of the 2013 updates now accumulates to 216.2 tonnes.
Central banks had in 2013 193.5 tonnes gold purchases that were kept hidden until now.
The gold/silver price ratio was thus driven up with fat central planners - "aid".
Meaning that the real hedgers against their fiat had a much smaller share in it.
It weren't speculators that moved from silver to gold.
In 2005, annual average price US $444.74, the central planners sold a record 663 tonnes gold
In 2013, annual average price US $1411.23, the central planners bought a record 625.5 tonnes gold.
SELL LOW
BUY HIGH
Guess who are the suckers that receive less fiat and less ounces? :D

SILVER
- A new Industrial component "Photovoltaic" was added, so the previous data from existing component "Other Industrial" is likely revised down in the same degree. Still, the differences are much bigger than Photovoltaic (plenty 25 Moz bigger drops)
- Net Government Sales 2005 increased 20.1 Moz
- Net Government Sales 2006 increased 16.4 Moz
- Jewelry 2013 increased 13.3 Moz
- Mine Production 2013 increased 15.7 Moz
- Industrial Fabrication 2013 increased 11.3 Moz
- Industrial Fabrication component Electrical & Electronics 2013 increased 32.2 Moz
(Other Industrial 2013 dropped 75.3 Moz, way more than the Photovoltaic of 59.9 Moz, so apparently some, initially as 'Other' declared, was later on reclassified as Electrical & Electronics).

Notice how the 2 biggest changes (though % wise much less than golds case) are again the central planners. Apparently in 2005-2006 they sold 36.5 Moz silver, to report it to Thomson Reuters in 2015. :D
 
i just..cant even...... the only figures i know is that 80% of the members and 100% of the lurkers have no idea what 100% of your posts are about.

few sentences in your own words what its about. pick 5 or so figures/examples from the entire post, maybe a link and thats it. it feels like your always trying to teach all of us DOS.
 
ego2spare said:
i just..cant even...... the only figures i know is that 80% of the members and 100% of the lurkers have no idea what 100% of your posts are about.

few sentences in your own words what its about. pick 5 or so figures/examples from the entire post, maybe a link and thats it. it feels like your always trying to teach all of us DOS.
Click to expand...
2013: ego2spare stated that he purchased 5 pigs of the rosy kind.
2015: ego2spare revises the 2013 figure to 10 pigs of the rosy kind.
The bastard bought 10 instead of 5 but didn't tell us back then the sneaky bastard! :D

http://thomsonreuters.com/en/about-us.html
About Us
Thomson Reuters is the world's leading source of intelligent information for businesses and professionals

http://nl.wikipedia.org/wiki/Thomson_Reuters

http://www.gold.org/supply-and-dema...+today:+Gold+Demand+Trends+First+Quarter+2015
http://www.gold.org/download/file/3755/gdt_q1_2015_tables.xlsx

http://www.silverinstitute.org/supply_demand.php

Gold and silver figures, revised (several times) later on.
(psst, that's the topics title).

I also don't bother that you "cant even". Should I?
 
ego2spare said:
it feels like your always trying to teach all of us DOS.
Click to expand...

lol gold

command-promt-commands-thumb.png


ipconfig.jpg
 
You must log in or register to reply here.
Back
Top